Multiple Time Series Modeling Using the SAS VARMAX Procedure by Anders Milhoj

Author:Anders Milhoj [Milhoj, Anders]
Language: eng
Format: azw3
Publisher: SAS Institute
Published: 2016-01-10T16:00:00+00:00

Program 11.2: Defining a Dummy Variable and a Preliminary Estimation of a VARMA(2,0) Model

DATA DUMMY;

SET SASMTS.EGG;

EUDUMMY=0;

IF YEAR(DATE)=1973 AND MONTH(DATE)=1 THEN EUDUMMY=1;

RUN;

PROC VARMAX DATA=DUMMY PRINT=ALL PLOTS=ALL ;

MODEL QEGG, PEGG, QTOT, PTOT=EUDUMMY/DIF=(QEGG(1) PEGG(1) QTOT(1) PTOT(1))

NSEASON=12 P=2 LAGMAX=25 XLAG=3 METHOD=ML;

ID DATE INTERVAL=MONTH;

RUN;

Estimation Results

The model contains many parameters, but the schematic presentation (Output 11.1) gives the overall picture. In Output 11.1, the periods (.) indicate insignificant parameters; the signs + and - indicate significant parameters at a 5% level. Many parameters are insignificant, which leads to the conclusion that the autoregressive part of the model is over-parameterized. The many stars (*) for the exogenous variable EUDUMMY denote parameters that are excluded from the model because the variable EUDUMMY in Program 11.2 affects only the last variable—that is, the total price variable PTOT. In the output, XL0, XL1, and so on are short for eXogenous at lags 0, 1, and so on.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.